Got departments buying ERP software as a service, rampant virtual-machine sprawl and a server room in the Idaho...
branch office with chronic uptime problems? Here's how to set up a centralized, uniform IT infrastructure.
IT management best practices suffered with the emergence of server rooms in branch offices, employee-led technology changes like bring your own device (BYOD) and departments buying their own data center equipment and applications. The result is a combination of multiple different facilities, each running to different service levels with poor overall systems utilization and a lack of overall systems availability. It's time to take that back.
What will help centralize IT?
Virtualization, touted as the ultimate answer to jumbled IT systems, made things worse -- the number of virtual machines (VMs) and idle applications spiraled out of control. Cloud computing -- another panacea -- required the organization to deal not only with its own issues around multiple facilities, but also with cloud providers outside their direct control.
Increased mobility of the workforce, both from those at home and those on the road, led to a need for continuous access to enterprise applications -- and also to a BYOD appetite for using apps from other environments.
For things to get better and not worse, complete a full audit of your own IT infrastructure. Identify every connected server within your network and every application running on them -- there are plenty of tools for this. Once you finish auditing, identify the physical location of each server. This may be slightly more difficult to do, but here's a suggestion: When you cannot pinpoint a server, deny its access to the main network. Within a few minutes, the help desk will receive a call from an unsatisfied user who will know exactly where that server lives.
After you have a physical asset map of your servers and you know which applications are running on them, identify any and all applications that are doing the same job. You may have three or four different customer relationship management systems, for example. Arrange to migrate off the nonstrategic systems as soon as possible. Then identify different instances running on the same application and consolidate these down as far as possible -- for example, there may be five different instances of the same enterprise resource planning (ERP) application in place.
Such functional redundancy is bad, not only for IT with the cost of the servers, operating system licenses, maintenance and power required to keep the mix of IT systems running, but also for the business. These systems will generally be running separately from each other, causing a distorted view of what is really happening.
Even with a consolidated environment, there will still be plenty of in-house applications that could be better sourced through a software as a service (SaaS) model. Software functionality that is highly dependent on domain expertise -- such as payroll and expense management -- should be outsourced to a third party that oversees the legal aspects. With SaaS-based solutions, you have no responsibility for the facility, hardware or software and can focus on the business' needs.
Dealing with your organization's overall IT platform with more controlled, yet flexible best practices is next.
The overall internal IT platform for the organization should be smaller than before. Consolidation, particularly when carried out with a fully planned virtualization strategy, can reduce IT equipment by as much as 80%. You now have control over where the equipment goes. Should it all be in an owned facility? Probably not.
There are problems in building and managing a highly flexible data center: Power distribution and cooling tend to be designed and implemented to meet specific needs. Consolidation can lead to the facility's power utilization effectiveness score growing rather than shrinking. Always-on uptime requirements mean multiple, different connections from the facility to the outside. The more that's kept within an owned facility, the more availability becomes an issue.
Treat data centers in a cascade design for the best IT management strategy. Some applications must remain in an owned facility because of a long-term investment in the application and/or IT equipment, fears over data security or other reasons. Many more applications can live in a colocation or hosting facility where a dedicated company provides and manages the building -- connectivity, cooling, power distribution and so on. You only have to manage the hardware and software in your space, or you can offload some of that to a hosting provider. If your business's IT needs grow, the facility owner can give you more space, power and cooling. If demand shrinks, you can negotiate for a smaller part of the facility.
For a cascaded data center design (consisting of an owned facility in conjunction with a colocation facility and/or public cloud functionality) to work, your IT team must have the right tools in place to manage the movement of workloads from one environment to another. It also requires effective monitoring of application performance with the capability to switch workloads around to maintain service levels. Getting it right will provide far greater flexibility at both the technical and business levels. IT departments should carry out a complete and effective audit as soon as possible to plan how to house and manage the IT platform in the future.